How to Use Statistics
The inferential statistical tests that we use in geography are based entirely on the concept of hypothesis testing. When a student comes to ask me what methods are appropriate for their particular problem, my first reply is always ‘what are your hypotheses’? Students often argue that they do not have any firm hypotheses, they just want to use some statistical methods to ‘explore their data’ and try to ‘see what it means’. They use terms like wanting to be ‘objective’. Personally, I think this is nonsense and that it is actually impossible to do any piece of research that involves gathering data unless you do have hypotheses; it is just that some people hide them away and refuse to acknowledge them. This is a very bad practice, not just because it prevents the use of inferential statistical methods, but because it also hinders progress.